Deep-learning model to improve histological grading and predict upstaging of atypical ductal hyperplasia / ductal carcinoma in situ on breast biopsy.

AIMS: Risk stratification of atypical ductal hyperplasia (ADH) and ductal carcinoma in situ (DCIS), diagnosed using breast biopsy, has great clinical significance. Clinical trials are currently exploring the possibility of active surveillance for low-risk lesions, whereas axillary lymph node staging may be considered during surgical planning for high-risk lesions. We aimed to develop a machine-learning algorithm based on whole-slide images of breast biopsy specimens and clinical information to predict the risk of upstaging to invasive breast cancer after wide excision. METHODS AND RESULTS: Patients diagnosed with ADH/DCIS on breast biopsy were included in this study, comprising 592 (740 slides) and 141 (198 slides) patients in the development and independent testing cohorts, respectively. Histological grading of the lesions was independently evaluated by two pathologists. Clinical information, including biopsy method, lesion size, and Breast Imaging Reporting and Data System (BI-RADS) classification of ultrasound and mammograms, were collected. Deep DCIS consisted of three deep neural networks to evaluate nuclear grade, necrosis, and stromal reactivity. Deep DCIS output comprised five parameters: total patches, lesion extent, Deep Grade, Deep Necrosis, and Deep Stroma. Deep DCIS highly correlated with the pathologists' evaluations of both slide- and patient-level labels. All five parameters of Deep DCIS were significantly associated with upstaging to invasive carcinoma in subsequent wide excisional specimens. Using multivariate logistic regression, Deep DCIS predicted upstaging to invasive carcinoma with an area under the curve (AUC) of 0.81, outperforming pathologists' evaluation (AUC, 0.71 and 0.69). After including clinical and hormone receptor status information, performance further improved (AUC, 0.87). This combined model retained its predictive power in two subgroup analyses: the first subgroup included unequivocal DCIS (excluding cases of ADH and DCIS suspicious for microinvasion) (AUC, 0.83), while the second excluded cases of high-grade DCIS (AUC, 0.81). The model was validated in an independent testing cohort (AUC, 0.81). CONCLUSION: This study demonstrated that deep-learning models can refine histological evaluation of ADH and DCIS on breast biopsies, which may help guide future treatment planning.

Extension of C. elegans lifespan using the ·NO-delivery dinitrosyl iron complexes.

The ubiquitous and emerging physiology function of endogenous nitric oxide in vascular, myocardial, immune, and neuronal systems prompts chemists to develop a prodrug for the controlled delivery of ·NO in vivo and for the translational biomedical application. Inspired by the discovery of natural [Fe(NO)2] motif, herein, we develop the synthetic dinitrosyl iron complexes (DNICs) [Fe2(µ-SR)2(NO)4] (1) as a universal platform for the O2-triggered release of ·NO, for the regulation of ·NO-release kinetics (half-life = 0.6-27.4 h), and for the activation of physiological function of ·NO. Using C. elegans as a model organism, the ·NO-delivery DNIC 1 regulates IIS signaling pathway, AMPK signaling pathway, and mitochondrial function pathway to extend the lifespan and to delay the aging process based on the lifespan analysis, SA-ßgal activity assay, and next-generation RNA sequencing analysis. This study unveils the anti-aging effect of ·NO and develops DNICs as a chemical biology probe for the continued discovery of unprecedented NO physiology.

Prediction of In-Hospital Cardiac Arrest in the Intensive Care Unit: Machine Learning-Based Multimodal Approach.

Background: Early identification of impending in-hospital cardiac arrest (IHCA) improves clinical outcomes but remains elusive for practicing clinicians. Objective: We aimed to develop a multimodal machine learning algorithm based on ensemble techniques to predict the occurrence of IHCA. Methods: Our model was developed by the Multiparameter Intelligent Monitoring of Intensive Care (MIMIC)-IV database and validated in the Electronic Intensive Care Unit Collaborative Research Database (eICU-CRD). Baseline features consisting of patient demographics, presenting illness, and comorbidities were collected to train a random forest model. Next, vital signs were extracted to train a long short-term memory model. A support vector machine algorithm then stacked the results to form the final prediction model. Results: Of 23,909 patients in the MIMIC-IV database and 10,049 patients in the eICU-CRD database, 452 and 85 patients, respectively, had IHCA. At 13 hours in advance of an IHCA event, our algorithm had already demonstrated an area under the receiver operating characteristic curve of 0.85 (95% CI 0.815-0.885) in the MIMIC-IV database. External validation with the eICU-CRD and National Taiwan University Hospital databases also presented satisfactory results, showing area under the receiver operating characteristic curve values of 0.81 (95% CI 0.763-0.851) and 0.945 (95% CI 0.934-0.956), respectively. Conclusions: Using only vital signs and information available in the electronic medical record, our model demonstrates it is possible to detect a trajectory of clinical deterioration up to 13 hours in advance. This predictive tool, which has undergone external validation, could forewarn and help clinicians identify patients in need of assessment to improve their overall prognosis.

A multiple phenotype imputation method for genetic diversity and core collection in Taiwanese vegetable soybean.

Establishment of vegetable soybean (edamame) [Glycine max (L.) Merr.] germplasms has been highly valued in Asia and the United States owing to the increasing market demand for edamame. The idea of core collection (CC) is to shorten the breeding program so as to improve the availability of germplasm resources. However, multidimensional phenotypes typically are highly correlated and have different levels of missing rate, often failing to capture the underlying pattern of germplasms and select CC precisely. These are commonly observed on correlated samples. To overcome such scenario, we introduced the "multiple imputation" (MI) method to iteratively impute missing phenotypes for 46 morphological traits and jointly analyzed high-dimensional imputed missing phenotypes (EC impu ) to explore population structure and relatedness among 200 Taiwanese vegetable soybean accessions. An advanced maximization strategy with a heuristic algorithm and PowerCore was used to evaluate the morphological diversity among the EC impu . In total, 36 accessions (denoted as CC impu ) were efficiently selected representing high diversity and the entire coverage of the EC impu . Only 4 (8.7%) traits showed slightly significant differences between the CC impu and EC impu . Compared to the EC impu , 96% traits retained all characteristics or had a slight diversity loss in the CC impu . The CC impu exhibited a small percentage of significant mean difference (4.51%), and large coincidence rate (98.1%), variable rate (138.76%), and coverage (close to 100%), indicating the representativeness of the EC impu . We noted that the CC impu outperformed the CC raw in evaluation properties, suggesting that the multiple phenotype imputation method has the potential to deal with missing phenotypes in correlated samples efficiently and reliably without re-phenotyping accessions. Our results illustrated a significant role of imputed missing phenotypes in support of the MI-based framework for plant-breeding programs.